A short blog post on the ongoing implications for a family home as it becomes reconstituted through the addition of data that is streamed from smart objects. Living in an environment that is equipped with an Internet-of-Things (IoT) involves the placement of multiple sensors that record change in conditions, in order to construct a simulacrum of the actual house from which to analyse and form understandings of behaviour, and in turn opportunities for connection. Through reviewing the imprint of their family, in the data gathered over time, occupants find that they are not only able to identify their routines, but also single out the routines of individual family members.
This post explores an author’s personal experience as he became aware of the activities of his partner, son and daughter. An awareness that offered insights that were previously forgotten such as the toilet habits of children who were once dependent on him to change nappies and supervise toilet training. These intimate endeavours are now revealed in patterns within data sets.
The post explores the implications of a ubiquitous domestic vision, as personal routines and habits that were previously hidden behind doors and walls become visible. The thoughts are a reflection upon the initial findings of the Hub-of-All-Things (HAT) project that involves the collection of domestic behaviours through sensors on objects in homes to uncover insights of into patterns of use and consumption.
Figure 1 – Feed from the toilet roll holder. Full toilet rolls can be identified as having the highest value up the Y axis, and significant use of the toilet roll can be identified in the drops in data.
Through the summer of 2014, my family was the subject of a study for the HAT project that involved attaching sensors to parts of the co-investigators’ homes. As a member of the design team within the project, Chris Barker (www.designinformatics.org) took on the task of developing a toilet roll holder that would record how much toilet paper was on the existing roll that was in use, and send the data to an online database.
The toilet roll holder was considered an interesting subject of design and development because it could provide (on the surface) an easy case study of an instrumented product that would provide the project participant who owned it, personal data about toilet roll consumption. The assumption was that this unique and very personal data would place the participant in a position of power in a new market – that of providing data about the use of toilet roll. Whilst this may sound trivial, toilet paper can be considered a staple requirement of many western homes and at present the business model for distribution is simply for manufacturers to pass them to shops for sale. The hypothesis for building the smart toilet roll holder was that knowing more about the consumption of toilet roll paper could inform the distribution, manufacture and purchase of toilet rolls (in a similar way to the Just-In-Time model developed by Toyota in the 1970s).
From the 5th of April to the present day, the amount of toilet roll in one toilet of the author’s house has been available on the internet (http://dev.modusdeus.com/toilet/vis/ see Figure 1. live feed above). The instrumented toilet roll holder was constructed using the Electric Imp cloud service and embedded hardware and software platform, featuring an infrared sensor to detect the amount of paper that was on the roll – see Figure 2. As more toilet paper was used, the distance between the paper and roll, which sat on a spindle, grew. Figure 2 demonstrates the use of toilet roll paper, with drops in data associated with the use of paper.
Living with the toilet roll holder and its data however proved to be far more disruptive than anticipated, and although it is early days before the research project understands how such personal data contributed to building a domestic market for selling data, the social impact was significant.
Originally identified by the research team as an easy ‘smart’ object to build compared to fridges and other domestic containers which hold multiple products, the toilet roll is at the centre of highly personal practices that take place behind locked doors. Through the online graph it is possible to clearly identify events that use significant amounts of toilet paper from which it is further possible to infer particular toilet activities (see Figure 3). Upon further analysis the graph revealed to the author and his family (all of whom have agreed to the study and to the sharing of the data), a series of likely events including cleaning up after cats, the running out of toilet paper, extra house guests and somebody having a runny nose.
Although the members of the family are quite happy to give the data away, the same data when correlated with other datasets could identify individuals and present a loss of privacy.
During the preliminary install of bathroom sensors in the author’s house, a wireless passive infrared sensor was placed in the bathroom to detect the presence of people (Figure 4). When the device detects movement, a small red LED lights on the surface of the object telling the occupant that they have been detected. Following ethical protocols during installation, the family was introduced to the sensor and their permission was sought to gather data. However when consulting the author’s daughter (8) she posed the question: “Can it see me on the toilet?”. An apparently simple question and one that the author could only truthfully answer as “yes”. However, the author proceeded to describe what the sensor could actually see, and once his daughter was happy that the device couldn’t identify her in person, concern for the device passed.
Nevertheless the disruptive nature of these interventions didn’t go away and once the first prototype toilet roll holder was installed the family experienced further surprises. On one occasion during July 2014, whilst the family was on holiday far away from the toilet, the software engineer who developed the device contacted the author over telephone to tell him of fresh activity on the server. Following a quick conversation with his partner, the author concluded that the change in toilet roll coincided with the fortnightly visit of the house cleaner.
Such an event highlights the complex market that surround objects that are connected to the internet and distribute data. Although consent to gather data was acquired for the family, this is not the case for visitors to the house and in particular of users of the toilet. In addition to the ethical consequences, the event also highlighted the function that connected objects can have beyond sensing the material that they are designed to. For the software engineer, the use of the toilet roll was an alert similar to that of a burglar alarm, and the data an indicator of an intruder.
The toilet roll holder and its connected database represent a suite of new technologies that are already available to buy. Whilst the brief summary of the social disruption that the toilet roll data has for a family is made clear due to the personal nature of toilet roll use, we should expect that all manner of conclusions may be drawn from even the most apparently benign sensor – from thermostat to toaster.
Living in a house that gains a shadow of itself in the form of` data sets had begun to change the way that the author understands his family. Not for a long time have I had to consider the toilet habits of my children or for that matter my wife. The last nappy I had to change was probably 5 years ago and soon after, the last of the nappy wipes was used up. Since then I have lost touch with when my children need the toilet, no longer do I have to smell them or ask them if they need to use the toilet. No longer do they even tell me when they need the toilet. The lack of this particular knowledge is welcome and I haven’t had to think about these particular practices for a long time. With a daughter of eight years of age and a son of 11, the lock on the toilet door has been in regular use for four years now and everybody who uses it is secure in the fact that they have achieved a level of privacy, and that no one else needs to be aware of what goes on behind the closed door. But more recently things have changed.
With an accurate toilet roll sensor I now know when my family has used the toilet. The door that was previously locked tight shut has now opened just a little. As visitors come to the house I am now obliged to advise them that the downstairs toilet roll holder is online and that if they would prefer to use an upstairs toilet they are quite welcome. I have even begun to wonder if I should redact particular activity on the online graph when guests use toilet paper, or perhaps I should return their data to them on a memory stick when they leave the house. Certainly the toilet door now requires terms and conditions that should be agreed upon before entering.
On the upside I am beginning to learn more about the family’s toilet paper habits and the software engineer who developed the technology is starting to explore opportunities for me to use the data in to inform living practices. The data in Figure 3 is beginning to offer information that is rich enough to begin programming alerts for me. For example, it appears that as the family runs out of toilet paper somebody places a half-used toilet roll onto the holder, perhaps a sign of desperation. Chris Barker (software engineer) is able to identify this shift in data and organise a Tweet to alert me to buy more toilet paper. Whilst a simple program for ordering toilet paper may not be so interesting, the important note is that the source of this data is not only personal to me, it is also owned by me. We built the toilet roll holder and I own the data. There are very few products or smart phone apps that I can say the same about. Usually I find myself agreeing to all manner of data agreements in order to get the ‘free’ software that is on offer. The toilet roll holder is then my first experience of producing data that I own and that I have the potential to begin to trade with. This awareness is the first step toward a significant turn in global economics – away from the established Push economy to a data-driven Pull economy.